“We get better results,” says Nara Logics’ CEO Jana Eggers when asked what separates the MIT-connected startup from their competitors. “On a fundamental level, we’ve proven that over and over again.” Eggers and CTO, Co-founder of Nara Logics, Nathan Wilson have encountered customers slogging through multiple AI solutions without achieving outcomes good enough to enter into production. The Nara Logics solution helps companies make better decisions by organizing siloed data to provide truly contextualized answers in an instant.
They’re pulling back the veil on AI, providing context so customers can understand the answers they are being offered, allowing them to respond and react with confidence. “The ability to have human control and human input is critical to what we do and how we operate,” explains Eggers. This truly symbiotic relationship between AI and human involvement consistently leads to better recommendations for personalization and decision support. “We’re getting 40 to 50 percent, sometimes 200 percent, better results than alternative solutions,” says Eggers. All while generating trust in the process.
At a time when enterprises are searching for opportunities to keep up with the data solutions curve, often getting bogged down along the way, Nara Logics has found a way to connect the dots between siloed data, present it in a manner that is understandable, and then tune that information so it is more accurate the next time around. Inspired by Wilson’s work in MIT’s Department of Brain and Cognitive Sciences, the Nara Logics Synaptic Intelligence Platform goes beyond the scope of other AI tools, including traditional neural networks.
Modeled loosely on the human brain—or, what Wilson refers to as “cognitive abstractions” thereof—traditional neural nets, while responsible for many of the best performing AI systems in the past decade, suffer from a variety of issues. For example, they require massive amounts of data and tend to rely heavily on pattern matching to predict future decisions. Meaning they are prone to define a pattern where one does not exist. But perhaps the most publicized reason that companies balk at utilizing machine learning models that employ neural net architectures is the fact that they operate using hidden layers or nodes, resulting in unexplainable outputs. Hence the common “black box” AI moniker.
However, the Nara Logics Synaptic Network suffers from none of these problems. Consider it Wilson’s Great Leap Forward. “We’ve developed a system that learns very quickly from a very small amount of data,” says Wilson. “It’s a brain you can put behind your interfaces so that your end users can get the right information at the right time, without having to change your interface, without having to change your data at the root,” says Wilson.
Nara Logics has found a foothold with companies in a wide range of industries, from retail and consumer packaged goods, to finance, process engineering, as well as the military and intelligence communities. Eggers notes that while they typically deal with companies in the global 200 or 500 arena, Nara Logics is more than capable of attending to the needs of smaller companies. They are all about going where they’re needed.
“The number one thing that we look for in a customer is that they are serious about their digital transformation,” says Eggers. That means helping organizations intent on utilizing their data to explore transformational change related to how they work, how they interact with customers, and how to improve operations. Their recent partnership with Proctor & Gamble highlights not only the power of the Nara Logics platform but also the philosophy of what Eggers and Wilson bring to the table.
When P&G wanted to differentiate Olay Skin Advisor in a marketplace filled with increasingly tech savvy customers looking for ever greater levels of personalization in their shopping experiences, they turned to Nara Logics. With Nara Logics powering their skincare recommendation tool, Olay doubled their ecommerce conversion rates. Because customers want accurate recommendations and they want to understand why a product is being recommended to them.
This philosophy of explainability extends to everyone on the ground using the Nara Logics platform. Consider the factory manager who is alerted to a problem on the production line or the financial analyst using AI to analyze how account holders spend and invest in order to customize advice. How do you engender trust in a recently sourced recommendation if you aren’t privy to the reasons behind that recommendation? The short answer is that you can’t.
“Every answer that we surface from our machine for decisions to help our end user is accompanied by a full back trace of the reasons why the machine thinks that’s a good answer,” says Wilson. “We think that’s important not just because we are employing it in industries where trust is paramount, but because it makes the machine smarter.”
As for companies questioning whether or not they need to be actively involved in AI, Eggers makes the point that if AI is going to be driving decisions, it is in essence a representation of company values. “AI really represents the values of who created it. If you’re going to have AI driving decisions in your company, you really need to be involved in what it is,” she says. Because data is the other key component along with the algorithm, Eggers believes that enterprises should be focused on how that data can be leveraged by AI for their goals—they must understand their data, how it is being used by the algorithm, and how it affects their goals.
“We fundamentally believe it’s important that companies be active in creating their own AI solutions,” says Eggers. And while that might sound self-serving from the CEO of a tech startup that provides a platform their customers can control, one gets the feeling that she isn’t selling a prepackaged line. In addition to boasting 25 years of technology and leadership experience that ranges from work with tiny startups to large enterprises like Intuit, Lycos, American Airlines, and Los Alamos National Laboratory, she’s a sought-after speaker, writer, and mentor on AI. She’s experienced, knowledgeable, and passionate about working with teams to define and deliver what customers want and need.
The result, aside from impressive numbers that prove the effectiveness of the Nara Logics platform in black and white, is a customer base that raves about working with Eggers and her team. “We’re very proud of how our customers feel about us. It means a lot to us,” says Eggers. “When our customers say we’re really easy to work with, part of it is because we realize it’s not just about the technology,” she says. While the technology is essential, the team at Nara Logics realizes their success depends very much on their ability to help an organization understand how to work with AI in order to be successful.
These days, Nara Logics has no shortage of customer interest, causing Eggers to reflect on the benefits of being associated with the Institute and programs like ILP and the Startup Exchange. But her appreciation goes beyond a sense of recognition. “Working with ILP provides us with a true sense of community,” she says. “It’s a place where startup founders can meet up to share their challenges and opportunities.” For Wilson, who has seen his research blossom from the lab to the real world, ILP plays an integral role in facilitating the MIT “mind and hand” ethos. “I think MIT is a special place—just look at the companies it creates. ILP makes sure those companies and their ground-breaking ideas get out into the real world and take root properly.”